Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.

5, October 2012
DOI : 10.5121/sipij.2012.3507 81

Incorporating Ridges with Minutiae for Improved
Fingerprint verification

Ms.M.Indra
#1
, Ms.S.Krishnaveni
#2
, Ms.P.Suriyapriya
#3

#1,#2,#3
Asst.Prof, Department of Information Technology,
Sethu Institute of Technology, Virudhunagar District,TamilNadu, India.
indrahere8@gmail.com,krish.jan1st@gmail.com,spria.p16@gmail.com


Abstract

Next to DNA, fingerprint is the unique feature which identifies the individual. Distortions and skin
deformations makes the fingerprint unreliable and it is difficult to match using minutiae alone. But when
ridge features are incorporated with minutiae features (minutiae type, orientation and position) more
topological information can be obtained. And also ridges are invariant to transformations such as rotation
and translation[1]. Ridge based coordinate system is used to extract the ridge features such as ridge
length, ridge count, ridge type and curvature direction in the skeletonized image. Breadth First Search is
used to traverse the graph formed using the minutiae as the node and the ridge vector formed using the
ridge features as the edge. The proposed ridge feature gives additional information for fingerprint
matching with little increment in template size and can be used along with the existing minutiae features to
increase the accuracy and robustness of fingerprint recognition systems.

Index Terms

Breadth first search, distortion, geometric transformation, ridge features, ridge-based coordinate system,
ridge patterns, ridge vector.


1. INTRODUCTION

FINGERPRINT recognition is one of the most adopted technique for user identification. This is
considered as a most reliable feature and the cost of implementing fingerprint recognition
methods is very less than other biometric features. It is used in many forensic and commercial
applications such as criminal investigation, electronic personal ID cards,etc. Although there is a
significant improvement in fingerprint recognition, some challenging tasks may degrade the
efficiency of fingerprint matching systems. Hence the fingerprint matching methods should be
implemented in such a way that it should overcome all difficulties while matching. Nonlinear
distortions, presented in touch-based fingerprint sensing, make fingerprint matching more
difficult. As shown in Fig. 1, even though these two fingerprint images are from the same
individual, the relative positions of the minutiae are very different due to skin distortions. This
distortion is an inevitable problem since it is usually associated with several parameters [6], [3],
including elasticity of skin, nonuniform pressure applied by the subject, different finger
placement with the sensor, etc.
Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
82



(a) (b)
Fig. 1. Example of skin distortions.
To deal with the distortions in fingerprint images and improve the matching performance, various
methods have been proposed by many researchers. These can be roughly classified into several
groups: modeling the distortion of fingerprints [6], [9]; detecting the distortions using special
hardware or video sequences allowing some amount of distortion in the minutiae matching stages
[8].

Luo et al. [8] used changeable tolerance boxes in the minutiae matching process. Starting from
the center of the fingerprint image, the size of the tolerance boxes is increased till the border of
the fingerprint area is reached. Although many fingerprint matching methods have been
developed to cope with distortions, most of them are minutiae-based. Thus, they cannot use
more topological information (such as ridge shape) covering the entire fingerprint image and the
limitation of information still exists. In addition, these methods use complex data structures and
many parameters for fingerprint matching. Accordingly, it is hard to understand and implement
these methods accurately. Considering the facts mentioned above, instead of developing complex
distortion models or elaborate minutiae alignment algorithms, a new and simple matching scheme
by incorporating conventional minutiae features and additional ridge features associated with
corresponding minutiae sets is proposed. To extract the ridge features, a ridge-based coordinate
system is also defined. The ridge features consist of four elements: ridge count (rc), ridge length
(rl), ridge curvature direction (rcd), and ridge type (rt). These features are invariant to any
geometric transformations (rotation, translation) of the fingerprints and concisely represent the
relationships between the minutiae since the maintenance of ridge structures is robust to
distortions.

2. FINGERPRINT PREPROCESSING AND RIDGE FEATURE EXTRACTION

A. Fingerprint Preprocessing

After preprocessing the image, the required fingerprint features for the proposed method were
extracted (see Fig. 2). Some additional procedures for quality estimation and circular variance
estimation is also introduced to improve the quality. Aftyer dividing the image into 8 8 pixel
Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
83
blocks, the mean and variance values of each block of the image are calculated to segment the
fingerprint regions in the image.

Adaptive binarization is applied which selects varying threshold based on the intensity of pixels
in each block. The orientation of ridges is estimated using the formula given below. Apply least
square approximation for all the pixels in each block.



Fig. 2. Overall Preprocessing Steps
(1)
where g
x
and g
y
gradients taken along x and y axis respectively. After finished with the estimation
of each block direction, those blocks without significant information on ridges and furrows are
discarded based on the following formula,

(2)

For each block in the image, the certainty level E is compared with the fixed threshold value.
Based on the certainty level some blocks are regarded as a background blocks which are
considered to have no significant information. These blocks have no information needed for
matching.

Histogram equalization is applied to enhance the image and to improve its contrast. When the
input is taken through the sensor, some region in the image becomes invisible or broken.
Histogram equalization method connects those false broken points. histeq() expands the pixel
values of the image. Fingerprint enhancement is done by dividing the image into small blocks (32
by 32 pixels) and perform the Fourier transform for the pixels of the image taken throughout all
the rows and columns of the input image Finally a skeletonized ridge image is obtained by the
Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
84
process of thinning and other morphological operations. The end and bifurcation minutiae
points are marked in the skeletonized image.

In order to avoid extracting false minutiae and to improve the confidence level and accuracy of
the extracted minutiae set, the quality estimation procedure is performed. The direction of each
block image of 16 by 16 pixels size is estimated by calculating the gradients along x and y axis.
Two dimensional filter is created using fspecial(sobel) where sobel is used to emphasize the
edges in the image.

The preprocessing stage alone is not enough to enhance the fingerprint image taken from the
sensor. For example, false ridge breaks and ridge cross-connections due to over inking are not
totally eliminated. These features result in the formation of spurious minutiae which will reduce
the efficiency and accuracy of the matching process. So in order to develop a accurate fingerprint
matching system, the spurious minutiae which are considered to be false minutiae have to be
removed before extracting the ridge features.

B. Ridge Feature Extraction

1) Proposed Ridge-Based Coordinate System

After performing the preprocessing steps, the skeletonized ridges and minutiae information is
obtained from the fingerprint image. Fingerprint Image is binarized to transform the 8-bit Gray
image into a binary image which has distinct values and colours for ridges and furrows. The
ridge coordinates are defined to extract ridge features between two minutiae. The ridge-based
coordinate system is defined by considering a minutia (called origin) and vertical and horizontal
axes starting from the origin minutia.

First, the vertical axis is defined by drawing a line passing through the origin and orthogonal to
the orientation of the origin. In addition, to define the sign of the vertical axis according to the
origin, the cross product between the orientation of the origin and the vector pointing from the
origin to the side of the vertical axis is calculated.



Fig. 3. Ridge Based Coordinate System
In the ridge-based coordinate system, the ridge features that describe the relationship between the
origin and an arbitrary minutiae is given as follows.

Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
85
(3)

where rc is the ridge count, rl is the ridge length, rcd is the ridge curvature direction and rt is the
ridge type. These four components form a ridge-based feature vector between two minutiae
and this feature vector is used in the matching process.

2) Ridge Feature Extraction

In the general ridge count methods [7], [10], the number of ridges that intersect the straight line
between two minutiae in the spatial domain is counted. However, when the ridge counting line is
parallel to the ridge structures, the line may meet the same ridge at one point, at more than two
points, or at no point, due to skin deformation Therefore, unlike existing ridge-counting methods,
here, the ridge count (rc) is calculated by counting the number of ridges along the vertical axis
until the axis meets the ridge attached to the neighboring minutia. The vertical axis is
perpendicular to the ridge structures. Thus, the counted numbers are less affected by skin
deformation than in the results of the general ridge counting methods. The ridge count feature is
more robust to skin deformation. Furthermore, to increase the discriminating power of the ridge
count (rc) feature, the direction of the ridge count line is also considered. The ridge count (rc) is
not always a positive number and the sign of the ridge count follows the sign of the vertical axis.
If two minutiae are directly connected by the same ridge, the ridge count would be zero.

The ridge length (rl) is the distance between the horizontal axis of the image and the axis of the
corresponding minutiae. The ridge length value also has a sign and follows the sign of the related
horizontal axis to improve the discriminating power. The probability distribution of the
absolute difference of the ridge length elements are mostly less than 16 pixels. Ridge length is
taken as the distance between two minutiae points in the input image. This provides more
discrimatory information and improves the accuracy since the distance between two points is
always constant in the case of misplacement of finger when compared to the template image. The
distance between two points is calculated using the Euclideans distance formula given below.

Distance = (1 2)

+(1 2)

(4)


Fig. 4. Ridge curvature direction. (a) Concave shape. (b) Convex shape.
To use more topology information in ridge patterns for matching, the ridge curvature direction is
also considered. As shown in Fig. 3, even though the ridge count and ridge length values are very
Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
86
similar, the shapes of the ridge patterns may be different [concave shapeFig. 3(a); convex
shapeFig. 3(b)].

The ridge curvature direction is defined as follows:
rcd = sign (

i
*
i-1
) (5)
where v
i
represents the i
th
vector between the sampling points along the horizontal axis from the
intersection of the vertical and horizontal axes to the minutia n and N represents the number
of sampling points.

Due to the feature extraction error, skin condition changes, and different finger pressures, end
points may appear as bifurcations and vice versa. Therefore, considering these facts and to further
improve the discriminating power of ridge features, the ridge type (rt) is used as one of the ridge
features instead of a minutia type. To determine the ridge type (rt), each minutia is classified as
an end point or a bifurcation. If a minutia is an end point, there is only one ridge belonging to the
minutia. If a minutia is a bifurcation, there are three ridges connected to the minutiae.

The overall procedure for extracting ridge features is as follows:

1) Perform preprocessing steps and extract a ridge image from a fingerprint.

2) Traverse the ridge-valley structures along the vertical axis from each minutia origin.

a) Whenever the vertical axis meets with the ridges attached to a minutia, then extract the
ridge features such as ridge count, ridge length, ridge curvature direction and ridge type
from the origin to the minutia and form a ridge feature vector between the origin and the
corresponding minutiae.

b) Keep traversing all the ridges until one of three terminating conditions is satisfied.

3) If all minutiae in the image gone through the above steps (1) and (2), terminate the procedure.
Otherwise, return to step (2) so that all minutiae can be marked as origin.

The termination conditions include the following three cases:

1) The vertical axis of each minutiae reaches a background region which has no needed
information in the fingerprint image.
2) The vertical axis reaches a poor quality region with very less minutiae and ridge points in
the fingerprint image.
3) The vertical axis reaches a high circular variance region in the fingerprint image.

3. FINGERPRINT MATCHING

The ridge feature vectors between the minutiae in the ridge coordinate system can be expressed as
a directional graph whose nodes are minutiae and whose edges are ridge feature vectors. The
graph matching methods are adopted to utilize the ridge feature vectors in fingerprint matching.
Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
87
Chikkerur et al. [12] proposed a graph-based fingerprint minutiae matching method in a
Euclidean space. They first defined the local neighborhood of each minutia, called K-plet, which
consists of the K nearest minutiae from a center minutia. The comparison of two Kplets is
performed by computing the distance between the two strings obtained by concatenating the K
neighboring minutiae, sorted by their radial distance with respect to the center minutia.
Neighborhoods are matched by dynamic programming and a match of local neighborhoods is
propagated with a breadth-first fashion. Thus, we apply this matching scheme to our ridge-based
coordinate system, since the ridge-based coordinate system can be represented as a graph and
each coordinate system makes a local neighborhood. While matching, the ridge maps of two
fingerprints are taken which provides information about ridge features. The similarity between
two ridge maps is calculated by choosing one minutiae from each image as origin. Initially the
alignment stage is performed. From the origin all possible ridges are taken. The minimum ridge
length is selected to compare the similarity certainity for two ridges. The ridge associated with
each minutia is represented by using its coordinate values of the points on the ridge. Throughout
the average inter-ridge length L, only a few sample points are taken. The overall flow of the
proposed fingerprint matching algorithm is as follows:

1) Extract a pair of ridge based coordinate system, one from the enrolled fingerprint image
and the other from the input fingerprint image. Match them using dynamic programming.
2) Select the top degree of matched ridge-based coordinate pairs.
3) For every initially matched pair, a breadth-first search is performed to detect the
matched ridge- based coordinate pairs. This is done incrementally until all the possible
matches are found.
4) Perform validity check and count the number of minutiae points which are matched by
using the above three steps.
5) Iterate steps 3) and 4) N times and return the maximum number of matched minutiae.
6) Compute the matching score.

Dynamic programming is applied to find the optimal solution in matching two string sequences in
the enrolled and input ridge-based coordinates. The ridge feature vectors in a ridge-based
coordinate system are arranged in the order of their ridge count feature component (rc), then the
order is invariant intrinsically. Therefore, the feature vectors in a ridge-based coordinate system
can be stored as the elements of an ordered sequence. Thus, all the enrolled and input ridge-based
coordinates are compared one by one and a similarity score is computed for the dynamic
programming.

The similarity score is calculated as follows:
(6)
where X is the absolute difference between two feature vectors, w
1
is the correctly matched class,
and w
2
is the incorrectly matched class. So the similarity of correlating the two ridges is derived
from:

S =
m
i=0
x
i
X
i
/[
m
i=0
x
i
2
X
i
2
]^0.5, (7)

Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
88
where (x
i~
x
n
) represents the set of minutiae in first ridge and (X
i~
X
N
) represents the set of
minutiae in the second ridge. If the similarity value is greater than the threshold, transfer all the
minutia in both fingerprints based on the reference pair of minutiae. translate and rotate all other
minutia with respect to the reference minutia according to the formula,

(8)

where (x,y,) is the parameters of the reference minutia, and translation matrix (TM) is given by,





The rotation angle is based on the minutiae direction obtained in the preprocessing stage.
For the ridge feature vector, the three feature elements (ridge count, ridge length, and ridge
curvature direction) are used to calculate the scores and the ridge type feature is used to check the
validity of the candidate pairs using the relative position and orientation of the minutiae. If the
relative position and orientation of the minutiae in the coordinate pair are also matched, then it
can be sure that these minutiae are correctly matched. Then the number of matched minutiae are
counted and stored. Finally, after the execution of the BFS procedure for every initial matched
pair, the maximum number of matched minutiae between two fingerprints is found. The final
match ratio for two fingerprints is the ratio of the number of minutiae in the input fingerprint
image over the template fingerprint image. The matching score is calculated as 100*ratio and it
ranges from 0 to 100. If the resultant score is larger than a pre-specified threshold, then the two
fingerprints are concluded to be from the same finger.

To compute the matching score, we must consider both the degree of overlap between two
impressions and the degree of similarity of the overlapped region. Thus, the matching score can
be computed as follows:

S
m
=

(10)
where L, N
1
, and N
2
are the number of matched minutiae, the number of minutiae in an input
image, and the number of minutiae in a template image, respectively. m
0
and n
0
are the number
of minutiae in the overlapping regions of the query and template images, respectively. The
overlapped regions are where two fingerprints intersect after the linear transformation (translation
and rotation) using the matched minutiae



xi_new
yi_new
i_new
|

\
|
|
|

xi x ( )
yi y ( )
i ( )

=TM *
Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
89
4. EXPERIMENTAL RESULTS AND ANALYSIS

The recognition performances of two algorithms (the conventional minutiae-based matching
method [13] and the proposed method) is compared. To demonstrate the effect of the proposed
ridge features more generally, the conventional minutiae-based method, which is based on
popular minutiae features such as minutiae position, minutiae orientation, and minutiae type [13]
is chosen instead of the state-of-the-art minutiae-based algorithms which use additional specific
matching techniques. The conventional method utilizes several reference points for local
alignment and an adaptive tolerance box is used to calculate the number of matched minutiae. For
genuine matches, each impression of each finger is compared with other impressions of the same
finger. It is shown in Fig. 5 that the genuine acceptance rate is higher in the proposed method
when compared to that of the conventional minutiae based method.



Fig. 5. ROC curve showing the accuracy of proposed method
5. CONCLUSION

In ridge based coordinate system, the ridge count (rc) is calculated by counting the number of
ridges crossing the vertical axis until the axis meets the ridge attached to the neighboring minutia.
The vertical axis is perpendicular to the ridge structures. Thus, the counted numbers are less
affected by skin deformation than in the results of the general ridge counting methods. Ridge
count feature is more robust to skin deformation. Furthermore, to increase the discriminating
power of the ridge count (rc) feature, the direction of the ridge count line is also considered. The
ridge length value also has a sign and follows the sign of the related horizontal axis to improve
the discriminating power. With this approach , no complex data structure is involved and no
minutiae alignment algorithm is used. The proposed ridge features are invariant to any transform,
thus they can be used in addition to conventional alignment-free features in the fingerprint
identification or cancellable fingerprint area and gives additional information for fingerprint
matching with little increment of template size. The proposed fingerprint matching method can
be applied to images with small foreground area and images with low quality which contains very
Signal & Image Processing : An International Journal (SIPIJ) Vol.3, No.5, October 2012
90
less number of minutiae points than required. This proposed fingerprint matching method can
also be enhanced by combining level 3 pores along with ridge features for improved accuracy.
Before extraction of the fingerprint features, the fingerprint is enhanced to improve the robustness
of the fingerprint. For this enhancement purpose, STFT (Short Time Fourier Transform
Analysis) can be used. If such algorithm is implemented, all the intrinsic properties of the
fingerprints such as the foreground region mask, local ridge orientation and local ridge frequency
can be estimated in a less time.

REFERENCES

[1] Heeseung Choi, Kyoungtaek Choi, and Jaihie Kim, Fingerprint Matching Incorporating Ridge
Features With Minutiae, IEEE Trans. Inf. Forensics Security, vol. 6, no. 2, pp. 338-345, Jun. 2011.
[2] D. Maltoni, D. Maio, A. K. Jain, and S. Prabhakar, Handbook of Fingerprint Recognition. New
York: Springer-Verlag, 2003.
[3] X. Chen, J. Tian, X. Yang, and Y. Zhang, An algorithm for distorted fingerprint matching based on
local triangle feature set, IEEE Trans. Inf. Forensics Security, vol. 1, no. 2, pp. 169177, Jun. 2006.
[4] X. P. Luo, J. Tian, and Y. Wu, A minutia matching algorithm in fingerprint verification, in Proc.
15th ICPR, Sep. 2000, vol. 4, pp. 833836.
[5] S.Chikkerur, A. N. Cartwright, and V. Govindaraju, K-plet and coupled BFS: A graph based
fingerprint representation and matching algorithm, Lecture Notes Comput. Sci., Adv. Biometrics,
vol. 3832, pp. 309315, 2005.
[6] A.Ross, S. Dass, and A. K. Jain, A deformable model for fingerprint matching, Pattern Recognit.,
vol. 38, no. 1, pp. 95103, 2005.
[7] X. Jiang and W. Y. Yau, Fingerprint minutiae matching based on the local and global structures,
in Proc. 15th Int. Conf. Pattern Recognition, Barcelona, Spain, Sep. 2000, vol. 2, pp. 1038104.
[8] N. K. Ratha and R. M. Bolle, Effect of controlled acquisition on fingerprint matching, in Proc. 14th
ICPR, 1998, vol. 2, pp. 16591661.
[9] Asker M. Bazen and Sabih H. Gerez, Thin-Plate Spline Modelling Of Elastic Deformations In
Fingerprints, Proc. 3rd IEEE Benelux Signal Processing Symposium (SPS-2002), Leuven, Belgium,
pp. S02-1 S02-4, March 21-22, 2002.
[10] Z. M. Kovacs-Vajna, A fingerprint verification system based on triangular matching and dynamic
time warping, IEEE Trans. Pattern Anal. Mach. Intell., vol. 22, no. 11, pp. 12661276, Nov. 2000..
[11] S. Lee, H. Choi, and J. Kim, Fingerprint quality index using gradient components, IEEE Trans. Inf.
Forensics Security, vol. 3, no. 4, pp. 792800, Dec. 2008.
[12] S. Chikkerur, A. N. Cartwright, and V. Govindaraju, K-plet and coupled BFS: A graph based
fingerprint representation and matching algorithm, Lecture Notes Comput. Sci., Adv. Biometrics,
vol. 3832, pp. 309315, 2005.
[13] D. Lee, K. Choi, and J.Kim, A robust fingerprint matching algorithm using local alignment, in
Proc. 16th Int. Conf. Pattern Recognition, Quebec City, Que., Canada, Aug. 2002, vol. 3, pp. 803
806.
[14] Sharat Chikkerur, Alexander N. Cartwright and Venu Govindaraju, Fingerprint enhancement using
STFT analysis, Pattern Recognition 40 (2007) 198211.